Syndrome Without Benign Intracranial Hypertension

J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.45.3.209 on 1 March 1982. Downloaded from

Journal ofNeurology, Neurosurgery, and Psychiatry 1982;45:209-216

Intracranial pressure in patients with the empty sella syndrome without benign intracranial hypertension

ANDREW H KAYE, BRIAN M TRESS, DAVID BROWNBILL, JOHN KING From the Department of Neurology and Neurosurgery, and the Department of Radiology, Royal Melbourne Hospital, Melbourne, Australia

SUMMARY The intracranial pressure was monitored continuously for at least 48 hours in five patients with empty sella syndrome, who did not have clinical benign intracranial hypertension (BIH). It has been suggested that the empty sella syndrome is a result of chronically elevated intracranial pressure in the presence of a congenitally deficient diaphragma sellae. However, whilst the intracranial pressure in two of the five patients was abnormally high, in three patients in whom it was monitored, the CSF pressure was normal. Although, these cases may represent "burnt out" forms of intracranial pressure problems, it might be that the normal pulsations of CSF are sufficient to produce the empty sella in the presence of a deficient diaphragma sellae.

"Empty sella" is a gross descriptive term introduced It has been suggested that the primary empty Protected by copyright. by Busch in 1951 to describe the appearance at sella is caused by a combination of a congenitally necropsy of the sella turcica when the diaphragma deficient diaphragma sellae and chronically elevated sellae is incomplete or forms only a small peripheral CSF pressure.5 11 12 rim.' In 1924, Schaeffer described the gross appear- There is a definite relationship between benign ance of the diaphragma sellae in 125 cases, and noted intracranial hypertension (BIH) and the primary that the diaphragm varied from being a complete empty sella syndrome'6 911-13 Although it has been roof which transmitted the infundibulum to, at the suggested that there must be a chronic elevation of other extreme, a small peripheral rim.2 In 40 of the intracranial pressure (ICP) in addition to a deficient 788 patients with no known pituitary disease studied diaphragma sellae to produce a prolapse of the by Busch (an incidence of 5 5 %), the diaphragma subarachnoid space into the sellae turcica,12 there sellae was a peripheral rim of tissue two millimetres has been no adequate measurement of the ICP of or less, with the pituitary gland flattened to the patients with the empty sella syndrome who do not bottom of the sella. have benign intracranial hypertension. The term "empty sella syndrome" originally This is a report of a series of five patients all with referred to the findings at surgical exploration in radiologically proven empty sella syndrome, who patients who had received radiation for an intrasellar did not have benign intracranial hypertension. In all http://jnnp.bmj.com/ tumour, and subsequently developed visual symp- patients the ICP was monitored continuously for at toms.3 Primary (idiopathic) and secondary forms of least 48 hours. the empty sella syndrome have been described.4 The primary empty sella has been defined as "that which Patients and methods did admit significant air at pneumoencephalo- graphy, in the absence of prior surgery or radiation All five patients were adult females who presented to the therapy".5 The role of a chronically elevated Royal Melbourne Hospital with a major complaint of cerebrospinal fluid (CSF) pressure in the patho- longstanding headaches. There was no significant neuro- on October 2, 2021 by guest. genesis of the so-called idiopathic empty sella logical deficit in any patient and all had normal visual has been discussed.6-10 fields on perimetry. A full clinical and laboratory endo- syndrome crine assessment was made in all patients and all patients had a normal serum level of prolactin. No significant Address for reprint requests: Andrew H Kaye, Department abnormality was found in any patient. All patients had of Neurosurgery, Radcliffe Infirmary, Oxford OX2 6HE, UK. neuroradiological investigations proving the existence of Received 28 January 1981 and in revised form 13 June 1981 an empty sella. The diagnosis of empty sella was based Accepted 25 October 1981 on a combination of an enlarged sella with characteristic 209 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.45.3.209 on 1 March 1982. Downloaded from

210 Kaye, Tress, Brownbill, King "symmetrical ballooning" or globular appearance on plain skull radiograph5 and the detection of CSF density within the sella by computed tomography (CT) performed with slice widths collimated to five millimetres and overlapped in both transverse axial and direct coronal planes. The diagnosis was confirmed in one case by pneumo- a) Case 1 encephalography and another by CT metrizamide cisternography.14 15 CT has been successful in diagnosing the presence of an empty sella in up to 100% of cases.14 16 17 False positive diagnoses have been produced by prolactin secreting tumours with necrotic centres,16 18 a possibility which was excluded in this series by pituitary function tests. The ICP was monitored continuously in all patients c) Case 3 d) Case 4 for a minimum of 48 hours by a subarachnoid screw inserted in the right frontal position (just in front of the coronal suture). This was connected by a 150 cm saline- filled tube to a Hewlett Packard transducer and then to a Hewlett Packard amplifier which was connected to a continuous paper recorder. A Hewlett Packard 1208C transducer was attached to a Hewlett Packard 78200 e) Case 5 monitor a series patient which houses 78205B amplifier. Fig 1 Tracings of the sella turcicas of each of the five This was attached to a Hewlett Packard 7803B monitor- cases, as seen in a standard lateral skull radiograph. to a scope and Rikadenki single channel recorder, In all cases the volume of the sella was increased. In Melbourne modified by the Royal Hospital Electronics particular the depth was increased, being in the range Protected by copyright. Department, for two channel multiplex operation. The 15-16 mm (upper level normal 14 mm). zero reference point for the transducer was taken as a point three centimetres anterior to the external auditory meatus, with the patient lying supine.19 The zero drift 'E and gain drift of the system was checked, and if necessary Z. corrected, at least every six hours. Very little correction C 0

was required at any time. Antibiotics (cloxicillin and c C:) ampicillin) were administered during the monitoring 40- a. 'o period. -4-44 Case I This 67-year-old obese woman, referred for investigation of the cause of two months' constant headache, had been previously reported by one ofthe authors20 (AHK). Neurological and general examination was normal. The plain skull radiographs revealed an enlarged 0 globular sella (fig la). CT showed normal size ventricles, 1 Tine (hours) 2 but also an enlarged sella with density consistent with that of CSF. A pneumoencephalogram showed air entering Fig 2 Intracranial pressure recording of Case 1, of and the showing a moderately raised baseline pressure with the anterior and mid portions the sella, http://jnnp.bmj.com/ pituitary flattened against the dorsum and posterior plateau waves up to 34 mm Hg. portion of the floor of the sella. During the follow-up examination three months after presentation, she devel- a right subfrontal craniotomy. At surgery an empty sella oped intermittent CSF rhinorrhoea. Continuous ICP was found and a fold of arachnoid was noted to extend monitoring revealed a generally raised baseline pressure down into and occupy the sella. A hole, present at the with intermittent high plateau waves up to 34 mm Hg junction of the floor with the anterior wall of the sella, (fig 2). On the basis of the elevated ICP and its presumed was packed with crushed muscle and covered with fascia causal relationship with the empty sella syndrome and lata.

CSF rhinorrhoea, it was decided to perform a lumbar She has subsequently been followed for two years and on October 2, 2021 by guest. peritoneal shunt as an initial procedure rather than a has had no further CSF rhinorrhoea. direct intracranial attack on the fistula itself. After operation, the patient was initially free of CSF rhinor- Case 2 This 70-year-old obese Italian lady presented rhoea and the ICP, measured again by the above tech- after 12 months of moderately severe continuous head- nique, demonstrated a normal baseline pressure and the ache. On examination there was no neurological abnor- elimination of wave forms. However, after 10 weeks there mality and endocrine studies were normal. Lumbar was a recurrence of the CSF rhinorrhoea. She was puncture pressure was 12 cm of CSF. A plain skull readmitted to hospital and the sella was explored through radiograph showed an enlarged globular sella (fig lb). J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.45.3.209 on 1 March 1982. Downloaded from

Intracranial pressure in patients with the empty sella syndrome without benign intracranial hypertension 211 CT using the above technique showed normal sized ventricles with an enlarged sella containing intrasellar .t- vio F4 attenuation values of CSF (fig 3). -v+!$ The ICP monitoring for 56 hours showed a slightly : elevated baseline pressure, with plateau waves rising to I E 30 mm Hg and Lundberg "B" waves21 up to 35 mm Hg E i i T i - -. '', ' (fig 4). These "B" waves were present during 20% of the a- monitoring time and there was no particular predilection a,_I-I*V.....- - 1 : ,., - for them to occur during the sleeping period. 1-t d kw - ,-. t -' Case 3 This 28-year-old obese female presented after 20 -'IJLL-.Ljl L.' 111;"Jul P,' some years of headaches. Neurological examination and kML-W: _lrt endocrine studies were normal. Plain skull radiographs 1~:F -11 2 showed the characteristically enlarged globular sella 1 2 (fig 1c), and CT, using the above technique, showed Time (hours) normal size ventricles with an enlarged sella containing Fig 4 ICP trace of Case 2, showing a slightly elevated CSF density material (fig 5). baseline pressure, with plateau waves rising to 30 mm The ICP was monitored for 48 hours. The baseline ICP Hg and Lundberg "B" waves up to 35 mm Hg. Protected by copyright. http://jnnp.bmj.com/

aLi. h. c (Ca%c 2. Over/appiulgl 5 /lilt/ Ihick IViIt'sIX /AtXX ny1ir/1of/Ifluidcu%iJII ithIIi ini on October 2, 2021 by guest. malerial "Itverchrmpin tlillX(/R{fif/ wfithsir1 r1is[ i J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.45.3.209 on 1 March 1982. Downloaded from

212 Kaye, Tress, Brownbill, King

Fig 5 a, b Ccase 3. Two 5 mm thick transverse axial scans through sella turcica, showing low density contents.

was within normal limits. Lundberg "B" waves occurred severe headache. There was no significant neurological for approximately 5 % of the monitoring time, particularly abnormality and endocrine studies were normal. during sleep, and rose to as high as 20 mm Hg (fig 6a and A plain skull radiograph showed the characteristic b). features of scaphocephaly. Convolutional impressions were marked particularly in the upper third of the vault,

Case 4 This 34-year-old rather thin female, with a past and the sella was symmetrically enlarged in a globular Protected by copyright. history of asthma and a hysterectomy for menstrual fashion (figs Id and 7). CT performed in transverse axial irregularity, presented following some years of moderately and direct coronal planes confirmed the presence of CSF density material in the sella (fig 8). The ICP was monitored for 48 hours. The baseline ICP pressure was within normal limits and there was no significant wave formation (fig 9). 20i (i c, ~~~~~~~u0L 5 obese female presented fol- I i .~0 Case This 41-year-old many years of constant headache. There was no E lowing neurological abnormality and a full endocrinological assessment was normal. A plain skull radiograph showed an enlarged globular

Time (hours)

0L u http://jnnp.bmj.com/

--o on October 2, 2021 by guest.

Time (minutes) Fig 6 a, b ICP trace of Case 3, showing normal baseline pressure (a) and Lundberg B waves Fig 7 Plain skull radiograph of Case 4, showing occurring during sleep (b), being present for marked convolutional impressions, particularly in the approximately 5 % of the monitoring time. upper third of the vault, and an enlarged globular sella. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.45.3.209 on 1 March 1982. Downloaded from lIltracrallijal prc.ssiire il paitielnt%s lit/i thc cimptl scllai sYnldronic it7tlout /)bcnign,}u iuitr-ticrciiliaitlhypcrtcisioni 21 3 Protected by copyright.

Fig 8 Case 4 (a, b, c) ()Oerlapping 5 n1un1 t/ick transverse axial Scans thlrioug/i sella tlecica a11d (d) 5 min thick coronal slice tlinaug/i sella, demonstorcting the sella to coantaill cerebro.spinalfluicd den sit, ;niatel ial. http://jnnp.bmj.com/ sella (fig le). CT in transverse axial and coronal planes, with and without intravenous contrast medium, showed c E O CSF density material within the sella (fig 10). CT per- O, 2 formed in the same planes after the introduction of 2 C intrathecal metrizamide showed a marked increase in the ;t t t e t - . density of the intrasellar contents, confirming the presence I of an empty sella (fig I 1). _EE The ICP was monitored continuously for 56 hours. The 0- baseline ICP was within normal limits and there was no on October 2, 2021 by guest. significant wave formation (fig 12). Il 1. ._LLu i, ,.1,1, !, i ., s 1 .. 1 I Ef Discussion 2 Time (hours) The most commonly accepted aetiologic hypothesis Fig 9 ICP trace of Case 4, demnionstratiuig n0ortm1al for the empty sella concerns transmission of elevated pr-essires. .... J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.45.3.209 on 1 March 1982. Downloaded from 214 Kaye, Tress, Brownbill, King Protected by copyright.

Fig 10 Case 5. Overlapping S mm thick scans in transverse axial (a, b) and coronal (c, d) planes, showing cerebrospinalfluid density material in sella. CSF pressure through a congenitally deficient sion of the subarachnoid space into the sella cavity

diaphragma sellae.5 1112 There is a definite relation- with enlargement and remodelling of the sella http://jnnp.bmj.com/ ship between benign intracranial hypertension and turcica and flattening of the pituitary contents development of the empty sella syndrome.6 11-13 against the floor.5 However, those patients with benign intracranial However, intermittent raised ICP cannot be hypertension form a minority in any series of patients adequately evaluated without using continuous ICP with empty sella syndrome, ranging from 8% of monitoring. cases,12 to 13 %5 and 158 %.6 Normal CSF pressure In this series of five patients with empty sella, none has been recorded in patients with empty of whom had benign intracranial hypertension on

sella.5 101122 However, in no study was continuous clinical grounds and in whom continuous ICP on October 2, 2021 by guest. ICP monitoring performed. The importance of recordings had been performed over at least a 48 continuous versus static measurements of ICP has hour period, a spectrum of ICP was seen from the been shown in normal pressure hydrocephalus, definitely abnormal, as in the patient (Case 1) who where intermittent high pressure peaks have been had marked wave formation, to those patients demonstrated.23 Neelon argues that longstanding (Cases 3, 4 and 5) who had normal ICP. Although mild elevation of CSF pressure in the setting of some Lundberg "B" wave activity has been reported incomplete sella diaphragm can bring about intru- as being normal,24 that seen in Case 2 was probably J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.45.3.209 on 1 March 1982. Downloaded from

Intracranial pressure in patients with the empty sella syndrome without benign intracranial hypertension 215 Protected by copyright.

Fig 11 Case 5. After introduction of intrathecal metrizamide the density of the sella contents is markedly increased, shown in both transverse axial (a, b) and coronal (c, d) http://jnnp.bmj.com/ planes.

hypertension. The recordings in those patients whose ICP is raised (Cases 1 and 2), help in the under-

cm20 _ XL sella I: -- --_a-:'- L standing of aetiology of the empty syndrome. Benign intracranial hypertension has been described on October 2, 2021 by guest. 10 ' in the absence of papilloedema,25-27 and Cases 1 and 2 may well represent examples of either benign 2 Time (hours) intracranial hypertension without papilloedema or a "forme fruste". The ICP traces in both cases are Fig 12 ICP time of Case 5, showing normal pressures. comparablenot onlywith those reported by Johnstone abnormal, in that it occurred for a large percentage and Paterson,28 in patients with papilloedema and of the monitoring time with a high amplitude. With benign intracranial hypertension, but also with those the exception of Case 1,20 ICP recordings have not cases of Spence et al,27 having the disorder without been previously reported in patients with empty sella papilloedema. Alternatively it has been suggested syndrome who do not have benign intracranial that obesity itself may cause a chronic elevation of J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.45.3.209 on 1 March 1982. Downloaded from

216 Kaye, Tress, Brownbill, King ICP.5 11 Whatever the cause of the pressure waves, and the intrasellar cistern. Radiology 1969;93:1085- the resulting raised ICP may be sufficient to prolapse 91. the arachnoid into the sella. 11 Kaufman B. The "empty" sella turcica-a manifesta- The normal ICP found in Cases 3, 4 and 5 does tion of the intrasellar subarachnoid space. Radiology ICP 1968;90:931-41. not necessarily negate the hypothesis of raised 12 Foley KM, Posner JB. Does pseudotumour cerebri contributing in the aetiology of empty sella syn- cause the empty sella syndrome? Neurology drome. The headache occurring in these patients is (Minneap) 1975;25:565-9. not readily explainable. In the patient (Case 4) who 13 Davis S, Tress B, King J. Primary empty sella syndrome had a scaphocephalic skull with marked copper and benign intracranial hypertension. Clinical and beating, it is interesting to speculate on the relevance experimental neurology. Proc Aust Assoc Neurol of the convolutional markings in the presence of the 1978 ;15:248-57. empty sella syndrome, and at the present time nor- 14 Bajraktari X, Bergstrom M, Brismar K, Goulatia R, mal ICP. These skull radiograph changes may Greitz T, Grepe A. Diagnosis of intrasellar cisternal ICP. The herniation (empty sella) by computer assisted tomo- represent evidence of previously raised graphy. J Comput Assist Tomogr 1977;1:105-16. ICP levels of the other two normal patients (Cases 3 15 Manelfe C, Pasquini V, Banks WO. Metrizamide and 5) may represent "burnt out" intracranial demonstration of the subarachnoid space surround- hypertension. That is, those patients with normal ing the optic nerve. J Comput Assist Tomogr 1978; ICP have had at some time in the development of the 215:545-7. empty sella a high CSF pressure. Alternatively, the 16 Rozaria R, Hammerschlag SB, Post KD, Wolpert SM, concept of the raised ICP being necessary for the Jackson I. Diagnosis of empty sella with C.T. scan. development of the empty sella syndrome may have Neuroradiology 1977;13 :85-8. Protected by copyright. to be reconsidered. It may be that the normal 17 Sage MR, Chan ES, Reilly PL. The clinical and are sufficient in the presence of radiological features of the empty sella syndrome. pulsations of CSF29 Clin Radiol 1980;31:513-9. an incomplete diaphragna sellae, to produce in some 1 Hsu TH, Shapiro JR. Hyperprolactinaemia associated patients the empty sella syndrome. with empty sella syndrome. JAMA 1979;235:2002-4. 19 Azevedo Filho HRC. Clinical applications of intra- The authors thank Mr David Wallace, Assistant cranial pressure monitoring. Oxford University, Neurosurgeon, Royal Melbourne Hospital for 1975;76-7. (MSc thesis.) permission to report Case 5. This work was under- 20 Davis S, Kaye AH. A dynamic pressure study of taken with help from a grant from the Victor spontaneous CSF rhinorrhoea in the empty sella Hurley Medical Foundation. syndrome. J Neurosurg 1980;52:103-5. 21 Lundberg N. Continuous recording and control of References intracranial pressure in neurosurgical practice. Acta Psychiatr Neurolog Scand 1960;36:(suppl 149): Busch W. Die morphologia der sella turcica and ihr 1-193. beziehungen zur hypophyse. Virchows Arch Pathol 22 Bemasconi V, Giovanelli MA, Papo I. Primary empty Anat 1951;320:437-58. sella. J Neurosurg 1972;36:157-61. 2 Schaeffer JP. Some points in the regional anatomy of 23 Symon L, Dorsch NWC, Stephens RJ. Pressure waves the optic pathways with especial reference to in so called low pressure hydrocephalus. Lancet http://jnnp.bmj.com/ tumours of hypophysis cerebri and resulting ocular 1972;2:291-2. changes. Anat Rec 1924;28 :243-79. 24 Martin G. Lundberg's B waves as a feature of normal 3 Colby MY, Kearns TP. Radiation therapy of pituitary intracranial pressure. Surg Neurol 1978;9(6):347-8. adenomas with associated visual impairment. Mayo 25 Cooper PR, Moody S, Skear F. Chronic monitoring of Clin Proc 1962;37:15-24. intracranial pressure using an in vivo calibrating 4 Weiss SR, Raskind R. Non neoplastic intrasellar cysts. sensor. Experience in patients with pseudotumour Inter Surg 1969;51:282-6. cerebri. Neurosurgery 1979;5 :666-70. Neelon FA, Goree JA, Lebovitz HE. The primary 26 Scanarini M, Mingrino S, d'Avella D, Della Corte V. empty sella. Clinical and radiographic characteristics Benign intracranial hypertension without papill- on October 2, 2021 by guest. and endocrine function. Medicine 1973;52:73-92. oedema. Case report. Neurosurgery 1979;5:376-7. 6 Berke JP, Buxton LF, Kokmen E. The "empty" sella. 27 Spence JD, Amacher AL, Willis NR. Benign intra- Neurology 1975 ;25:1137-43. cranial hypertension without papilloedema. Role of 7 Garcia-Uria J, Carrillo R, Serrano P et al. Empty sella 24 hours cerebro spinal fluid pressure monitoring and rhinorrhoea. J Neurosurg 1979;50:466-71. in diagnosis and management. Neurosurgery 1980; 8 Mortara R, Norrell H. Consequences of a deficient 7(4) :326-36. sella diaphragma. J Neurosurg 1970;32:565-73. 28 Johnstone I, Paterson A. Benign intracranial hyper- 9 Weisberg LA, Zimmerman EA, Frantz AG. Diagnosis tension II CSF pressure and circulation. Brain 1974; and evaluation of patients with an enlarged sella 97:301-12. turcica. Am J Med 1976;61:590-6. 29 Du Bourlay GH. Pulsatile movements of CSF path- 1.0Zatz LM, Janon EA, Newton TH. The enlarged sella ways. Br J Radiol 1966;39 :255-62. Book reviews 763 plasma as tools for obtaining biochemical Correction Book reviews and pharmacokinetic data in neuroleptic In the article "Intracranial pressure in therapy. They suggest some relationship patients with the empty sella syndrome exists between chlorpromazine treatment without benign intracranial hypertension" and clinical effect, at least in the early Kaye, Tress, Brownbill, King, J Neurol Psychotropic Drugs. Plasma Concentration phases of treatment. Fulton and others Neurosurg Psychiatry 1981;45:209-16 and Clinical Response. Edited by Graham contribute extensively on the phar- figure 10(d) was incorrect. The correct D Burrows and Trevor R Norman. (Pp macokinetic of benzodiazepines. Justifi- figure 10(d) is reproduced here. 544; SFr. 150.) New York: Marcel Dekker ably, diazepam is given pride of place but Inc, 1981. further sections deal with other individual benzodiazepine compounds. Sedatives and Correlation of drug effect with plasma hypnotics are dealt with by Wade but levels is of use in many disease states. One again the conclusion is that plasma con- of the most difficult areas, however, is that centrations of hypnosedatives correlate of psychiatric illness. The variety of means poorly with their effects, both at dosages of assessing or classifying mental illness, used clinically and in overdosage. Eadie the poor understanding of underlying demonstrates the value of anticonvulsant pathology and the changes in brain func- drug plasma levels but concludes that the tion that occur, explain why this area has patient's clinical state should always be the not seen more widespread implementation final criterion in management and that of plasma level monitoring. The volume plasma anticonvulsant levels should be edited by Burroughs and Norman, how- seen only as a means to an end. The book is ever, is a significant contribution to the lit- finished by an overview by Hol- erature on this topic. It is a comprehensive admirably volume dealing with many aspects of drug lister who concludes that there is little monitoring in the psychiatric field. The ini- doubt that monitoring plasma levels of of lithium and of anticonvulsants has made tial chapter by Lang on the mechanism such treatme-nt both safer and more action of psychotropic drugs, provides an effective. However, he is less certain of the interesting introduction to the pharmacol- role of tricyclic antidepressant monitoring ogy of these compounds. Two interesting and argues that there is little to favour chapters follow, on methods for the meas- monitoring of plasma concentrations of urement of psychotropic drugs-anti- antipsychotic drugs. Hollister concludes depressants, antipsychotics and anti- that anti-anxiety and hypnotic drugs, which anxiety agents-which deal comprehen- generally have a large margin of safety and sively with all modern techniques. The suitable clinical criteria of response, should chapter by Graham shows application of not be monitored. I find this overall to be a these techniques to the measurement of the remarkably good book that is thoroughly pharmacokinetics of tricyclic anti- recommended. I found much of interest depressants. This is followed by an excel- and much new information and the editors lent chapter by Burroughs and Norman must be congratulated for the production dealing with the relationship between of such a volume. I would imagine, how- plasma levels and clinical response in the ever, that at 150 Swiss francs many poten- anti-depressant field. The authors conclude tial be frightened away. that no consistent relationship exists bet- customers will ween plasma levels and clinical response PG JENNER for tiicyclic anti-depressants and that routine monitoring of plasma levels is not Prevention and Treatment of Depression. warranted. This contrasts with the meas- Thomas A Ban, Rene Gonzalez, Assen S urement of serum lithium levels dealt with Jablensky, Norman A Sartorius and Felix by Annitto and Gershon where only be- E Vartanian. (Pp 300; £19.50.) Baltimore, nefit is derived from determining correct University Park Press. Distributor: MTP therapeutic dosage and prevention in toxic- Press Ltd. Lancaster, 1981. ity. Excellent chapters also follow on antipsychotic agents, Curry dealing with chlor- This is the proceedings of a conference on promazine in considerable depth, Sakalus depression organised by WHO and held in and Traficante reporting on fluphenazine, Washington in June 1980. Despite its title Crammer on thioridazine and Evans on its 31 chapters embrace most aspects of butyrophenones. It must be said, however, depressive illness-epidemiology, that the overall conclusion from these in- classification, symptomatology, the role of depth appraisals is that little relationship biological and psychological factors in exists between efficacy and plasma or aetiology, various forms of treatment and blood antipsychotic levels. Sedvall and models of care delivery from various parts Grimm contribute on sampling CSF and of the world. The proceedings of confer-